Add basic vector support for MySQL 9.x (vitessio#16464)

Signed-off-by: Dirkjan Bussink <[email protected]>

go/mysql/binlog/constants.go

@@ -79,6 +79,9 @@ const (
 	// TypeTime2 is MYSQL_TYPE_TIME2
 	TypeTime2 = 19
 
+	// TypeVector is MYSQL_TYPE_VECTOR
+	TypeVector = 242
+
 	// TypeJSON is MYSQL_TYPE_JSON
 	TypeJSON = 245
 

go/mysql/binlog/rbr.go

@@ -111,7 +111,7 @@ func CellLength(data []byte, pos int, typ byte, metadata uint16) (int, error) {
 		return intg0*4 + dig2bytes[intg0x] + frac0*4 + dig2bytes[frac0x], nil
 	case TypeEnum, TypeSet:
 		return int(metadata & 0xff), nil
-	case TypeJSON, TypeTinyBlob, TypeMediumBlob, TypeLongBlob, TypeBlob, TypeGeometry:
+	case TypeJSON, TypeTinyBlob, TypeMediumBlob, TypeLongBlob, TypeBlob, TypeGeometry, TypeVector:
 		// Of the Blobs, only TypeBlob is used in binary logs,
 		// but supports others just in case.
 		switch metadata {
@@ -652,8 +652,8 @@ func CellValue(data []byte, pos int, typ byte, metadata uint16, field *querypb.F
 		return sqltypes.MakeTrusted(querypb.Type_SET,
 			data[pos:pos+l]), l, nil
 
-	case TypeJSON, TypeTinyBlob, TypeMediumBlob, TypeLongBlob, TypeBlob:
-		// Only TypeBlob is used in binary logs,
+	case TypeJSON, TypeTinyBlob, TypeMediumBlob, TypeLongBlob, TypeBlob, TypeVector:
+		// Only TypeBlob and TypeVector is used in binary logs,
 		// but supports others just in case.
 		l := 0
 		switch metadata {

go/mysql/binlog_event_rbr.go

@@ -43,6 +43,7 @@ const (
 	tableMapEnumAndSetDefaultCharset
 	tableMapEnumAndSetColumnCharset
 	tableMapColumnVisibility
+	tableMapVectorDimensionality
 )
 
 // This byte in the optional metadata indicates that we should
@@ -136,7 +137,7 @@ func metadataLength(typ byte) int {
 		// No data here.
 		return 0
 
-	case binlog.TypeFloat, binlog.TypeDouble, binlog.TypeTimestamp2, binlog.TypeDateTime2, binlog.TypeTime2, binlog.TypeJSON, binlog.TypeTinyBlob, binlog.TypeMediumBlob, binlog.TypeLongBlob, binlog.TypeBlob, binlog.TypeGeometry:
+	case binlog.TypeFloat, binlog.TypeDouble, binlog.TypeTimestamp2, binlog.TypeDateTime2, binlog.TypeTime2, binlog.TypeJSON, binlog.TypeTinyBlob, binlog.TypeMediumBlob, binlog.TypeLongBlob, binlog.TypeBlob, binlog.TypeGeometry, binlog.TypeVector:
 		// One byte.
 		return 1
 
@@ -172,7 +173,7 @@ func metadataRead(data []byte, pos int, typ byte) (uint16, int, error) {
 		// No data here.
 		return 0, pos, nil
 
-	case binlog.TypeFloat, binlog.TypeDouble, binlog.TypeTimestamp2, binlog.TypeDateTime2, binlog.TypeTime2, binlog.TypeJSON, binlog.TypeTinyBlob, binlog.TypeMediumBlob, binlog.TypeLongBlob, binlog.TypeBlob, binlog.TypeGeometry:
+	case binlog.TypeFloat, binlog.TypeDouble, binlog.TypeTimestamp2, binlog.TypeDateTime2, binlog.TypeTime2, binlog.TypeJSON, binlog.TypeTinyBlob, binlog.TypeMediumBlob, binlog.TypeLongBlob, binlog.TypeBlob, binlog.TypeGeometry, binlog.TypeVector:
 		// One byte.
 		return uint16(data[pos]), pos + 1, nil
 
@@ -198,7 +199,7 @@ func metadataWrite(data []byte, pos int, typ byte, value uint16) int {
 		// No data here.
 		return pos
 
-	case binlog.TypeFloat, binlog.TypeDouble, binlog.TypeTimestamp2, binlog.TypeDateTime2, binlog.TypeTime2, binlog.TypeJSON, binlog.TypeTinyBlob, binlog.TypeMediumBlob, binlog.TypeLongBlob, binlog.TypeBlob, binlog.TypeGeometry:
+	case binlog.TypeFloat, binlog.TypeDouble, binlog.TypeTimestamp2, binlog.TypeDateTime2, binlog.TypeTime2, binlog.TypeJSON, binlog.TypeTinyBlob, binlog.TypeMediumBlob, binlog.TypeLongBlob, binlog.TypeBlob, binlog.TypeGeometry, binlog.TypeVector:
 		// One byte.
 		data[pos] = byte(value)
 		return pos + 1

go/mysql/query.go

@@ -891,7 +891,7 @@ func (c *Conn) parseStmtArgs(data []byte, typ querypb.Type, pos int) (sqltypes.V
 			return sqltypes.NULL, 0, false
 		}
 	case sqltypes.Decimal, sqltypes.Text, sqltypes.Blob, sqltypes.VarChar, sqltypes.VarBinary, sqltypes.Year, sqltypes.Char,
-		sqltypes.Bit, sqltypes.Enum, sqltypes.Set, sqltypes.Geometry, sqltypes.Binary, sqltypes.TypeJSON:
+		sqltypes.Bit, sqltypes.Enum, sqltypes.Set, sqltypes.Geometry, sqltypes.Binary, sqltypes.TypeJSON, sqltypes.Vector:
 		val, pos, ok := readLenEncStringAsBytesCopy(data, pos)
 		return sqltypes.MakeTrusted(sqltypes.VarBinary, val), pos, ok
 	default:

go/mysql/query_test.go

@@ -460,6 +460,7 @@ func TestQueries(t *testing.T) {
 			// Skip TUPLE, not possible in Result.
 			{Name: "Type_GEOMETRY ", Type: querypb.Type_GEOMETRY, Charset: collations.CollationBinaryID, Flags: uint32(querypb.MySqlFlag_BINARY_FLAG | querypb.MySqlFlag_BLOB_FLAG)},
 			{Name: "Type_JSON     ", Type: querypb.Type_JSON, Charset: collations.CollationUtf8mb4ID},
+			{Name: "Type_VECTOR   ", Type: querypb.Type_VECTOR, Charset: collations.CollationBinaryID},
 		},
 		Rows: [][]sqltypes.Value{
 			{
@@ -492,6 +493,7 @@ func TestQueries(t *testing.T) {
 				sqltypes.MakeTrusted(querypb.Type_SET, []byte("Type_SET")),
 				sqltypes.MakeTrusted(querypb.Type_GEOMETRY, []byte("Type_GEOMETRY")),
 				sqltypes.MakeTrusted(querypb.Type_JSON, []byte("Type_JSON")),
+				sqltypes.MakeTrusted(querypb.Type_VECTOR, []byte("Type_VECTOR")),
 			},
 			{
 				sqltypes.NULL,
@@ -523,6 +525,7 @@ func TestQueries(t *testing.T) {
 				sqltypes.NULL,
 				sqltypes.NULL,
 				sqltypes.NULL,
+				sqltypes.NULL,
 			},
 		},
 	})

go/mysql/sqlerror/constants.go

@@ -125,6 +125,8 @@ const (
 	ERDupIndex                      = ErrorCode(1831)
 	ERInnodbReadOnly                = ErrorCode(1874)
 
+	ERVectorConversion = ErrorCode(6138)
+
 	// already exists
 	ERDbCreateExists = ErrorCode(1007)
 	ERTableExists    = ErrorCode(1050)

go/mysql/sqlerror/sql_error.go

@@ -245,6 +245,7 @@ var stateToMysqlCode = map[vterrors.State]mysqlCode{
 	vterrors.KillDeniedError:              {num: ERKillDenied, state: SSUnknownSQLState},
 	vterrors.BadNullError:                 {num: ERBadNullError, state: SSConstraintViolation},
 	vterrors.InvalidGroupFuncUse:          {num: ERInvalidGroupFuncUse, state: SSUnknownSQLState},
+	vterrors.VectorConversion:             {num: ERVectorConversion, state: SSUnknownSQLState},
 }
 
 func getStateToMySQLState(state vterrors.State) mysqlCode {

go/sqltypes/type.go

@@ -186,6 +186,7 @@ const (
 	HexVal     = querypb.Type_HEXVAL
 	Tuple      = querypb.Type_TUPLE
 	BitNum     = querypb.Type_BITNUM
+	Vector     = querypb.Type_VECTOR
 )
 
 // bit-shift the mysql flags by two byte so we
@@ -219,6 +220,7 @@ var mysqlToType = map[byte]querypb.Type{
 	17:  Timestamp,
 	18:  Datetime,
 	19:  Time,
+	242: Vector,
 	245: TypeJSON,
 	246: Decimal,
 	247: Enum,
@@ -331,6 +333,7 @@ var typeToMySQL = map[querypb.Type]struct {
 	Datetime:  {typ: 12, flags: mysqlBinary},
 	Year:      {typ: 13, flags: mysqlUnsigned},
 	Bit:       {typ: 16, flags: mysqlUnsigned},
+	Vector:    {typ: 242},
 	TypeJSON:  {typ: 245},
 	Decimal:   {typ: 246},
 	Text:      {typ: 252},

go/vt/sqlparser/ast_funcs.go

@@ -293,6 +293,8 @@ func SQLTypeToQueryType(typeName string, unsigned bool) querypb.Type {
 		return sqltypes.Set
 	case JSON:
 		return sqltypes.TypeJSON
+	case VECTOR:
+		return sqltypes.Vector
 	case GEOMETRY:
 		return sqltypes.Geometry
 	case POINT:

go/vt/sqlparser/keywords.go

@@ -752,6 +752,7 @@ var keywords = []keyword{
 	{"varcharacter", UNUSED},
 	{"variance", VARIANCE},
 	{"varying", UNUSED},
+	{"vector", VECTOR},
 	{"vexplain", VEXPLAIN},
 	{"vgtid_executed", VGTID_EXECUTED},
 	{"virtual", VIRTUAL},

go/vt/sqlparser/parse_test.go

@@ -1228,6 +1228,9 @@ var (
 		input: "select /* bool in order by */ * from t order by a is null or b asc",
 	}, {
 		input: "select /* string in case statement */ if(max(case a when 'foo' then 1 else 0 end) = 1, 'foo', 'bar') as foobar from t",
+	}, {
+		input:  "select 1 as vector",
+		output: "select 1 as `vector` from dual",
 	}, {
 		input:  "/*!show databases*/",
 		output: "show databases",
@@ -5924,6 +5927,10 @@ partition by range (YEAR(purchased)) subpartition by hash (TO_DAYS(purchased))
 			input:  "create table t (id int, s varchar(255) default 'foo\"bar')",
 			output: "create table t (\n\tid int,\n\ts varchar(255) default 'foo\"bar'\n)",
 		},
+		{
+			input:  "create table t (id int, vec VECTOR(4))",
+			output: "create table t (\n\tid int,\n\tvec VECTOR(4)\n)",
+		},
 	}
 	parser := NewTestParser()
 	for _, test := range createTableQueries {